Planetary and Space Science 77 (2013) 64–73
Contents lists available at SciVerse ScienceDirect
Planetary and Space Science
journal homepage: www.elsevier.com/locate/pss
Magnetospheric ion sputtering and water ice grain size at Europa
T.A. Cassidy a,n, C.P. Paranicas b, J.H. Shirley c, J.B. Dalton III c, B.D. Teolis d, R.E. Johnson e,
L. Kamp c, A.R. Hendrix c
a
Laboratory for Atmospheric and Space Physics, University of Colorado Boulder, 1234 Discovery Dr., Boulder, CO 80303, USA
Applied Physics Laboratory, Johns Hopkins University, 11100 Johns Hopkins Road, Laurel, MD 20723, USA
c
Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Road, Pasadena, CA 91109, USA
d
Southwest Research Institute, Space Science and Engineering Division, 6220 Culebra Rd., San Antonio, TX 78238, USA
e
University of Virginia, Charlottesville, VA 22904, USA
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 January 2012
Received in revised form
24 June 2012
Accepted 9 July 2012
Available online 27 July 2012
We present the first calculation of Europa’s sputtering (ion erosion) rate as a function of position on
Europa’s surface. We find a global sputtering rate of 2 1027 H2O s 1, some of which leaves the surface in
the form of O2 and H2. The calculated O2 production rate is 1 1026 O2 s 1, H2 production is twice that
value. The total sputtering rate (including all species) peaks at the trailing hemisphere apex and decreases
to about 1/3rd of the peak value at the leading hemisphere apex. O2 and H2 sputtering, by contrast, is
confined almost entirely to the trailing hemisphere. Most sputtering is done by energetic sulfur ions (100s
of keV to MeV), but most of the O2 and H2 production is done by cold oxygen ions (temperature 100 eV,
total energy 500 eV). As a part of the sputtering rate calculation we compared experimental sputtering
yields with analytic estimates. We found that the experimental data are well approximated by the
expressions of Famá et al. for ions with energies less than 100 keV (Famá, M., Shi, J., Baragiola, R.A., 2008.
Sputtering of ice by low-energy ions. Surf. Sci. 602, 156–161), while the expressions from Johnson et al. fit
the data best at higher energies (Johnson, R.E., Burger, M.H., Cassidy, T.A., Leblanc, F., Marconi, M., Smyth,
W.H., 2009. Composition and Detection of Europa’s Sputter-Induced Atmosphere, in: Pappalardo, R.T.,
McKinnon, W.B., Khurana, K.K. (Eds.), Europa. University of Arizona Press, Tucson.). We compare the
calculated sputtering rate with estimates of water ice regolith grain size as estimated from Galileo NearInfrared Mapping Spectrometer (NIMS) data, and find that they are strongly correlated as previously
suggested by Clark et al. (Clark, R.N., Fanale, F.P., Zent, A.P., 1983. Frost grain size metamorphism:
Implications for remote sensing of planetary surfaces. Icarus 56, 233–245.). The mechanism responsible for
the sputtering rate/grain size link is uncertain. We also report a surface composition estimate using NIMS
data from an area on the trailing hemisphere apex. We find a high abundance of sulfuric acid hydrate and
radiation-resistant hydrated salts along with large water ice regolith grains, all of which are consistent with
the high levels of magnetospheric bombardment at the trailing apex.
& 2012 Elsevier Ltd. All rights reserved.
Keywords:
Europa
Water Ice Sputtering Yields
Galileo NIMS Data Analysis
1. Introduction
The Galilean satellites are exposed to intense space weathering
by ions and electrons in Jupiter’s magnetosphere. This has been
invoked to explain large-scale features in the moons’ reflectance
spectra. The trailing hemisphere, which is upstream with respect
to magnetospheric flow, is darker and redder than the leading
hemisphere (e.g., Pospieszalska and Johnson, 1989) and has more
sulfur compounds owing to the Iogenic sulfur-rich plasma bombarding the moon (e.g., Hendrix et al., 2011). Relating these
n
Corresponding author. Tel.: þ1 434 806 9880.
E-mail address: [email protected] (T.A. Cassidy).
0032-0633/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.pss.2012.07.008
features to magnetospheric bombardment requires the calculation of ion flux as a function of position on the surface.
In this paper we describe calculations of ion flux and sputtering (erosion) rate. Sputtering is a particularly robust process at
Europa’s surface owing to the bombardment of energetic heavy
ions (Cooper et al., 2001). Because sputtering selectively removes
smaller regolith grains, we compare our calculated sputtering
pattern with grain size as determined by recent analyses of
Galileo NIMS (Near Infrared Mapping Spectrometer) data (see
companion paper by Dalton et al. (2012)). We find that the
calculated sputtering rate is correlated with regolith grain size;
large calculated sputtering rates correspond to large grain sizes.
We also present, for the first time, surface composition and grain
size information for a location near Europa’s trailing side apex at
01, 2701W.
T.A. Cassidy et al. / Planetary and Space Science 77 (2013) 64–73
Pospieszalska and Johnson (1989) made the most recent
estimates of the ion bombardment pattern. Since then the Galileo
mission has provided new information about Europa’s plasma
environment (Paranicas et al., 2009; Kivelson et al., 2009). Though
they only considered the flux of sulfur ions and did not calculate
the sputtering rate, other papers have taken their fluxes as a
proxy for sputtering rate (e.g., Tiscareno and Geissler, 2003;
Cassidy et al., 2008). We show here that ion flux and sputtering
rate are quite distinct.
In this paper we cover a variety of topics with the goal of
calculating ion flux maps and sputtering rates. Sputtering an ice
surface produces mostly H2O with a fraction of O2 and H2, but the
latter are important in that they are the primary suppliers of
Europa’s atmosphere and neutral gas cloud (or ‘torus’).
2. Calculation details
2.1. Model description
The model we used to calculate ion flux is similar to
Pospieszalska and Johnson (1989), we ‘trace’ ion trajectories as
they gyrate around Jupiter’s magnetic field lines and ‘drift’
relative to Europa. The ion motion is treated as a superposition
of three motions (see Fig. 1): (1) A circular motion perpendicular
to the magnetic field line with a radius called the gyroradius and
speed given by vsinðaÞ, where v is the ion speed and a is called the
pitch angle, the angle of the ion’s velocity relative to the field line.
(2) Motion parallel to the field line with speed vcosðaÞ. (3) A third
component is a motion perpendicular to the field line called the
guiding-center drift. The predominant part of this drift is called
the E_ B_ drift, which in this case provides ‘corotation’, as if the
magnetic field line it gyrated about rotated with Jupiter. This
corotation drift actually lags slightly behind Jupiter’s rotation, we
used a value of 76 km s 1(Kivelson et al., 2009). We also included
the gradient-curvature drift (Thomsen and Van Allen, 1980).
This was found to be much less important than the corotation
E_ B_ drift for the ions that sputter Europa’s surface and so we
will not discuss it further. This component of the drift is, however,
important for the relativistic electrons that bombard Europa
(Paranicas et al., 2009; Dalton et al., this issue). These are not
discussed here because electrons have small sputtering yields
compared to ions (see Section 2.3).
We assume that Jupiter’s magnetic field, in the vicinity of the
region we are simulating, is anti-parallel to Europa’s spin axis, as
in Pospieszalska and Johnson (1989). Jupiter’s tilted dipole actually results in a time-varying, slightly tilted magnetic field at
Europa’s orbit, but Europa’s interior conductivity (due to the
presence of an ocean) cancels out the time-varying component
of the field near Europa’s surface, leaving just the anti-parallel
component (Zimmer et al., 2000). Except for this consideration,
we treat Europa as electromagnetically inert.
65
One difference between Pospieszalska and Johnson (1989) and
the current model is that we trace the ions backward in time from
Europa’s surface. This is more computationally efficient in that we
only calculate ion trajectories that intersect Europa’s surface.
The method works as follows: for a given location on Europa’s
surface, ion incidence direction and energy, the ion’s motion is
traced backward in time. If the ion’s reverse motion intersects the
sphere defining Europa’s surface, then that trajectory is forbidden; ion flux to that location has been blocked by Europa’s
surface. If the ion travels sufficiently far ‘upstream’ of Europa,
then that ion’s trajectory is allowed and its flux is added to that
location’s total, the ion flux being defined by
,
^
ðv nÞdf
ðE, OÞDEDO
ð1Þ
Where f(E,O)DEDO is the ion energy and angular distribution
measured near Europa’s orbit, but far from Europa itself. It is also
called the phase space density, and is described in the next
section. This is multiplied by the energy and solid angle incre,
^ is the
ments used by the program, d is the ion density, and ðv nÞ
ion speed relative to the surface dotted into the surface normal.
This is another improvement over Pospieszalska and Johnson
(1989), who did not quantify the flux, only bombardment
patterns.
The description above is equivalent to employing Liouville’s
theorem, which states that the phase space density f is conserved
along a dynamical trajectory. This is a rough approximation for
Europa as f is known to change through mechanisms such as pitch
angle scattering (Williams and Mauk, 1997). Another approximation is that we treat Europa as electromagnetically inert. However
electrical currents induced in Europa’s ionosphere divert plasma
around Europa to an uncertain extent (Saur et al., 1998).
Such diversion would have a small effect on the energetic ions,
but could have dramatic effect on the ‘cold’ ions.
2.2. Plasma flux parameters
The phase space density f(E,O) and density d are given in several
recent reviews. The distribution function is broken up into two
different ion ‘populations’, reflecting the different instruments used
to characterize the populations. The distribution function is a sum of
these two populations. The ‘cold’ or ‘thermal’ population is described
by Kivelson et al. (2009), it is composed of oxygen ( 110 cm 3) and
sulfur (20 cm 3) and the energy distribution f is roughly a
Maxwellian with temperature 100 eV. There is also a 20 eV cold
hydrogen population (Paterson et al., 1999).
The much more energetic population is described by Mauk
et al. (2004) and Paranicas et al. (2009). We refer to this as the
‘hot’ population. The hot f is described by a power law extending
from 50 keV to 50 MeV, with most of the flux coming from ion
with energies less than 1 MeV. We used the parameters listed in
Paranicas et al. (2009), Table 1 for H þ , On þ , and Sn þ , calculating
densities from the fluxes reported there. The densities for the hot
Fig. 1. Schematic of ion motion perpendicular to Jupiter’s magnetic field line (pointing into the page). There is also motion parallel to the field.
66
T.A. Cassidy et al. / Planetary and Space Science 77 (2013) 64–73
Table 1
List of NIMS observation locations and estimated H2O regolith grain sizes. The average ice grain size from the linear mixture model is listed first, followed by the output of
the model. 14ENSUCOMP01 consists of multiple observations, so the model output is not given. In that case the solutions consisted almost entirely of 75 and 100 mm
grains. The non-ice portion of the surface consists of the various hydrates and a neutral dark absorber.
Observation
W. Longitude
Latitude
Grain Size
Symbol on Fig. 7
14ENSUCOMP01 (Shirley et al., 2010)
14ENSUCOMP03 (Dalton et al., 2012)
15ENSUCOMP01 Dalton et al. (2012)
17ENSUCOMP02 Dalton et al. (2012)
G1ENNHILAT01 (this paper)
1701–1851
2031
1141
1201
2711
261–291 South
19.51 North
7.31 North
631 South
131 North
85–100 mm
250 mm, 34 % 250 mm, 76 % non-ice
55 mm, 67% 50 mm, 15% 75 mm, 18 % non-ice
80 mm, 65% 75 mm, 16% 100 mm, 19 % non-ice
930 mm, 2.5% 750 mm, 6.3% 1000 mm, 92 % non-ice
Plus Signs
Circle
Asterisk
Triangle
Diamond
As mentioned above, Jupiter’s magnetic field is slightly tilted
relative to Europa’s orbit plane. As a result, Europa’s plasma
environment changes as the dipole spins. To account for this we
used averaged plasma parameters. The parameters we just
described apply to the ‘centrifugal equator’, the equilibrium
surface at which the cold plasma density maximizes (at Europa’s
distance, it lies between the planet’s magnetic equator and the
rotational equator), but Europa moves above and below this plane
by 1 Jupiter radius (RJ). The cold plasma decreases in density as
exp( (z/H)2) above the centrifugal equator with a scale height of
H 1 RJ (Bagenal, 1994), resulting in an average cold ion density of
0.75 times the centrifugal value above. The hot ions do not
decrease much in density above the centrifugal equator owing
to their uniform PAD (Roederer, 1970).
2.3. Sputtering yields
Fig. 2. (a) (top): Schematic of ion motion, looking down onto Europa’s North pole.
The gyroradius is to scale for a 1 MeV Sþþion with a 301 pitch angle. (b) (bottom):
Schematic of ion motion, looking from the side. This schematic demonstrates the
importance of motion perpendicular to the field line. The slant of the ion motion is
determined by its motion parallel to the magnetic field (vector at left) and the
corotation drift (rightward) motion. Slant is not to scale.
population are much lower than the cold population, about
0.4 cm 3 for Sn þ , 0.2 cm 3 for On þ , and 0.1 cm 3 for H þ .
The angular portion of f(E,O) is known as the pitch angle
distribution (PAD). We used isotropic PADs for both the hot and
cold populations. Mauk et al. (2004) wrote that the hot population
PAD in the vicinity of Europa’s orbit is nearly uniform (to within
25%). As for the cold ions, the PAD does not matter. This population is so cold that its thermal speed does not affect the flux onto
Europa’s surface, only the drift speed matters (see Fig. 2).
Both of these populations exhibit some temporal variability.
Kivelson et al. (2009) estimated that the cold ion density varies by
a factor of two, and Paranicas et al. (2009) estimated that the hot
ion flux also varies by about a factor of two.
The charge state of the ions is an important parameter as it helps
to determine the gyroradius of the ions (Fig. 2, top). This is a poorly
constrained quantity for the hot ions. Lagg et al. (2003) estimated
that the oxygen and sulfur ions are multiply charged, but tend
toward lower charge states over time by capturing electrons from
neutral gas orbiting Jupiter. We chose a charge state of 2þ for the
oxygen and 3þ sulfur ions, following Cooper et al. (2001) (and
references therein). We do not expect the charge state to affect the
sputtering yield; electrons are stripped from and captured by ions as
they travel through a solid so that incident ions quickly lose their
original charge state upon entering the surface.
Another quantity we need is the sputtering yield, the number
of H2O molecules ejected per incident ion (including other
molecules produced during bombardment, mainly O2 and H2).
The water ice sputtering yield is an experimentally determined
quantity that has been fit with various functional forms, including
dependencies on ion energy, ion mass, nuclear charge, water ice
temperature and incident angle. We found that we needed two
different functional forms to fit the available experimental data.
One approximation from Famá et al. (2008) is appropriate for low
(o100 keV) energy ions. This is the upper limit for applicability
specified in their paper. The other expression, from Johnson et al.
(2009), matches the data above 100 keV. These two fits and
available experimental data are shown in Fig. 3. Note that some
sputtering yields from Fig. 3 were corrected for non-normal ion
incidence and high temperature using the expressions from Famá
et al. (2008).1 This temperature dependence is important for
Europa, we explicitly include this effect in our modeling with
Eq. (2) below.
We included this temperature dependence in our calculated
sputtering yield by estimating the surface temperature with a
numerical thermal model (Spencer et al., 1989),2 using the albedo
and thermal inertia values from Spencer et al. (1999) as input.
This was implemented by calculating, at each latitude, the diurnal
average of the temperature factor mentioned above
YðTÞ ¼ YðT ¼ 0Þð1 þ 220 Exp½0:06 eV=kTÞ
ð2Þ
where Y is the sputtering yield. The expressions above apply to
the total sputtering yield, including all sputter products.
A sputtering yield of 1 means that the given ion ejects, on average,
the mass equivalent of one H2O molecule, though some portion
of the ejecta is actually O2 and H2 (more this below).
This temperature enhancement factor is about 2 at Europa’s subsolar point ( 130 K) and about 1 on the nightside equator ( 80 K).
1
2
These expressions are not unique to Fama’s paper, see, e.g., Johnson, 1990.
Code is available here: http://www.boulder.swri.edu/~spencer/thermprojrs/.
T.A. Cassidy et al. / Planetary and Space Science 77 (2013) 64–73
67
50.0
1.00
0.50
10.0
5.0
0.10
0.05
1.0
0.5
103
104
105
106
1000
103
104
105
106
1000
100
100
10
10
1
103
104
105
106
107
104
5000
104
1000
500
1000
105
106
107
100
100
50
10
1
10
104
105
106
107
104
105
106
107
Fig. 3. Compilation of sputtering yield data and theory. Experimental data points best match the Fama et al. curve (blue) below 100 keV and match the Johnson curve
(red) above 100 keV. There are no available experimental results for sulfur ion sputtering, but the calculated sputtering yield is shown here because sulfur is the most
important sputtering agent at Europa. The data were compiled from various experimental sputtering papers by R. E. Johnson and M. Liu at http://people.virginia.edu/~rej/
h2o.html. Note that the cold ions hit Europa at 500 eV (oxygen) or 1000 eV (sulfur) while the hot ions have a wide range of energies (100s of keV to MeV). (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Europa’s surface appears to be covered in a porous regolith
(Domingue and Verbiscer, 1997), and this affects the sputtering
yield. Most molecules sputtered from within a regolith actually
hit other grains, rather than directly leaving the surface. This can
lower the sputtering yield by around 75% if the sputter products
‘stick’ or freeze to the grains. At the same time the regolith
enhances the sputtering yield as ions cross the almost randomly
oriented grain surfaces with a variety of incidence angles, and
non-normal incidences increase the sputtering yield (e.g.,
Johnson, 1990). The two effects largely cancel each other out,
and we ignore them here (Cassidy and Johnson, 2005).
The surfaces of regolith grains are themselves are not likely to
be smooth at the atomic scale, but one computational study
(Rodriguez-Nieva et al., 2011) suggested that nm scale porosity
should not affect the sputtering yields.
Finally, the sputtering (ion erosion) rate is found by multiplying the calculated differential ion flux (cm 2s 1eV 1) by
the appropriate sputtering yield and integrating. As previously
noted, photons and electrons also sputter water ice, but have
much smaller sputtering yields, and so make only a small
contribution to the sputtering rate (Cooper et al., 2001; Plainaki
et al., 2010).
3. NIMS observations of H2O ice grain sizes
Below we relate the sputtering rate with H2O grain size. In this
section we describe the observations and grain size derivation.
Shirley et al. (2010) and Dalton et al. (2012) provide detailed
composition and H2O grain size information for multiple widely
spaced locations on Europa, based on an ongoing reanalysis of NIMS
data. They have measured the cryogenic spectra of hydrated
materials (e.g., H2SO4nH2O, Na2SO4nH2O) present on Europa’s
surface. They estimate Europa’s surface composition with a spectral
model that adds these reference spectra linearly, adjusting the
composition until it finds the best match to the NIMS spectra.
The model only estimates regolith grain size for the pure water
ice component of the surface, and so the grain size discussed here
is for the water ice component of the surface only. The water ice
reference spectra used for the mixture model were generated
with a radiative transfer model for a discrete set of grain sizes
between 25 and 5000 mm.
For the present study we employ the published data from
Shirley et al. (2010) and Dalton et al. (2012) (Table 1 and Fig. 4).
The observation footprints are shown in Figs. 4 and 7. We only use
spectral solutions from a single kind of geological terrain to avoid
comparison between areas of radically different composition or
geological history. We chose the ridged plains as they have been
exposed to Europa’s radiation environment for the longest time,
and their surface properties and compositions should reflect an
equilibrium between endogenic and exogenic influences.
While it may appear that the number of data points employed
for the study is small, it should be emphasized that each of the
spectra employed was obtained by averaging of dozens to
hundreds of carefully selected and processed individual NIMS
spectra. These are optimal conditions for the preservation of the
68
T.A. Cassidy et al. / Planetary and Space Science 77 (2013) 64–73
Fig. 4. Global image mosaic views of Europa with locations of the NIMS observations employed in this study. Left: Trailing hemisphere perspective Right: Leading
hemisphere perspective.
Fig. 5. Regional context image showing the source area (triangle) for the trailing
hemisphere apex location analyzed here. The triangular selection area was chosen
to avoid data drop-outs (due to incomplete wavelength coverage) that are present
in the NIMS dataset. The area shown represents the northwestern corner of the
G1ENNHILAT01 observation sub-area shown in Fig. 4. This study area was chosen
because it exhibits a higher abundance of water ice than is found for adjacent
areas to the south and east. The underlying basemap is available from the USGS
Flagstaff Planetary Geologic Mapping website (http://astrogeology.usgs.gov/Pro
jects/PlanetaryMapping/).
subtle signatures of the various ice grain sizes actually present on
the surfaces imaged.
In this paper we describe new results (including water ice
abundance and grain size) for a key location near Europa’s trailing
apex (observation G1ENNHILAT01 in Fig. 4). NIMS imaged a large
portion of Europa’s anti-Jovian hemisphere at moderate (38 km/
pixel) resolution during the first orbital encounter (G1) of the
Galileo Mission. The observation G1ENNHILAT01 spans latitudes
from below the equator to the northern polar regions, with trailing
side apex locations imaged at an oblique viewing angle near the
western limb.
The previous studies in Table 1 had high enough resolution to
permit the selection of ridged plains observations, whereas the
trailing apex observation reported here covers a broad area and
includes more than one kind of geological terrain. This exception
is due to the lack of higher-resolution coverage, preventing the
certain identification of geological terrain. A subsection of that
trailing apex observation was selected for analysis here, shown in
Fig. 5. It was chosen for its relatively high abundance of water
ice compared to other areas in the observation. We extracted an
area-average spectrum from the 66 NIMS pixels falling within the
area indicated on Fig. 5. This spectrum and the accompanying
spectral model solution are shown in Fig. 6.
The area selected for analysis (Fig. 5) includes a portion of
Conamara Chaos, a severely disrupted region characterized by
large (several km) tilted and rotated ice blocks separated by rough
hummocky matrix materials. The linear mixture model, however,
only estimates grain size for the pure water ice component of the
mixture, making it unlikely that this non-icy chaos terrain is
influencing the grain size estimate.
The presence of relatively large-grained ice in the mixture
model solution is a key result for the present investigation.
Ice grains with 1 mm and 750 mm diameters are not seen in any
of the other locations thus far analyzed (Dalton, 2007; Shirley
et al., 2010;, Dalton et al., 2012).
The 65% sulfuric acid hydrate abundance of Fig. 6 agrees well
with the 62% abundance obtained by Dalton (2007), who employed
an average spectrum for 18 contiguous visibly dark pixels from
G1ENNHILAT01. No water ice was called for in the prior solution,
which did not overlap with the area selection made here. The
present solution shows moderate abundances of hexahydrite and
epsomite, magnesium sulfate hydrates which are known to be quite
stable under irradiation (McCord et al., 2001). In general, the high
abundance of sulfuric acid hydrate, the presence of only the most
stable of the hydrated salts, and the large grain sizes are all
consistent with the expected effects of high levels of magnetospheric
bombardment (Clark et al., 1983; Carlson et al., 1999; McCord et al.,
2001; Dalton, 2007; Shirley et al., 2010; Dalton et al., 2012).
The accuracy of the grain size determinations is a complicated
matter. The sensitivity of the linear mixture modeling algorithms
to variations in water ice grain sizes is discussed in some detail in
Appendix C of Shirley et al. (2010). There it was shown that
single-step increases and decreases of water ice grain size give
rise to diagnostic differences in the spectra, particularly near
wavelengths of 2.05 and 2.25 mm. The mixture modeling algorithms are sensitive to these differences, returning significantly
larger w-squared values (indicating poorer fits) for cases where an
incorrect grain size was artificially specified. While we are unable
to rigorously quantify the error bars for the size determination,
the algorithm is clearly capable of distinguishing between the
spectra representing each of the grain size dimensions we have
employed. Further, as also shown in the cited paper, abundance
differences of Z3% of any component also yields noticeable
degradation of the w-squared values. Thus we believe a 3% error
bar (1-sigma) is appropriate for the stated percent abundances of
the single grain size components (50 mm, 75 mm, etc).
T.A. Cassidy et al. / Planetary and Space Science 77 (2013) 64–73
69
Fig. 6. Reflectance spectrum and linear mixture modeling results for Conamara Chaos and vicinity (Fig. 5). The extremely distorted water ice absorption bands at 1.5 and
2.0 mm are indicative of a high abundance of hydrated species (including hydrated sulfuric acid) in this location. 1 s error bars are provided. (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)
Sputtering Rate (cm-2 s-1)
1010
5x109
0
Fig. 7. Calculated sputtering (ion erosion) rate on Europa’s surface. Symbols show approximate locations where NIMS data has been analyzed. Table 1 gives the estimated
regolith grain sizes at these locations. Grain sizes and sputtering rates are shown in Fig. 10. Note on the right-hand side that the sputtering yield increases with latitude on
the leading hemisphere. We will show below that this is principally an effect of the preferential bombardment of the polar regions by hot ions. (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)
While the difference between 25 and 50 mm grain sizes may not
seem particularly important, one must at the same time recognize
that one of these is twice as large as the other; for spectroscopic
investigations, and considerations involving photon path lengths, this
difference is both easily detectable and quite significant.
In reality the imaged surfaces must be expected to exhibit a
continuum of ice grain sizes. We believe it is a strength of our method
that reference spectra representing multiple grain sizes may be
linearly combined in precise proportions to optimize the match
between observed and model spectra. While the combinations we
have obtained are not unique, we are unaware of any other technique
that is capable of improving on the estimates presented here.
4. Results
4.1. Europa sputtering rates
Fig. 7 and Table 2 show the results of our ion tracing model
calculations. Overall, there is about a factor of 3 difference
between the peak sputtering rates at the leading apex (901W
Longitude) and trailing apex (2701W Longitude). The largest
sputtering agent, as measured by the global sputtering rate, is
the hot sulfur ion population (Table 2), although cold oxygen is
important right at the trailing apex. Fig. 7 also shows the locations
where regolith grain sizes have been estimated using the Galileo
NIMS observations.
The globally-averaged sputtering rate is 7 109 cm 2 s 1.
This number is similar to earlier estimates that assumed uniform
ion flux. Cooper et al. (2001) estimated a factor of 4 lower
sputtering rate while Paranicas et al. (2002), surprisingly, calculated an identical number. This sputtering rate corresponds to an
erosion rate of 70 m per 109 years. Given that 1/3 of sputtered
H2O is has enough energy to escape Europa’s gravity, this
corresponds to a loss of 20 m per 109 years.
To better show the effect of ion energy on the sputtering map,
Fig. 8 shows sulfur ion flux maps for a single energy ion
population and a uniform PAD. At 102 eV, an energy typical of
the cold population, sulfur ions have very small gyroradii and the
ion speed (25 km s 1) is small compared to the corotation drift
speed (76 km s 1). This situation is shown schematically in Fig. 2
(ion trajectory labeled ‘cold’). The corotation drift carries the ions
70
T.A. Cassidy et al. / Planetary and Space Science 77 (2013) 64–73
Table 2
Sputtering rates and ion fluxes for each ion population. Note that the peak sputtering rate occurs on the trailing hemisphere apex
(2701W), except for the hot hydrogen, where the peak sputtering rate is at the poles.
Species
Total sputtering rate
Total ion flux
Peak sputtering rate
O2 sputtering rate
Cold Sn þ
Hot Sn þ
Cold On þ
Hot On þ
Cold Hn þ
Hot Hn þ
Total
8 1025 s 1
1 1027 s 1
3 1026 s 1
2 1026 s 1
9 1023 s 1
3 1024 s 1
2 1027 s 1
9 1024 s 1
1 1024 s 1
4 1025 s 1
6 1023 s 1
1 1024 s 1
2 1024 s 1
1 109 cm 2 s 1
4 109 cm 2 s 1
4 109 cm 2 s 1
8 108 cm 2 s 1
1 107 cm 2 s 1
2 107 cm 2 s 1
3 1025 s 1
2 1025 s 1
8 1025 s 1
6 1024 s 1
1 1023 s 1
3 1024 s 1
1 1026 s 1
100 eV
105 eV
103 eV
104 eV
106 eV
107 eV
Fig. 8. Sþþ flux patterns as a function of energy. Each map was produced by using a single energy ion energy with an isotropic pitch angle distribution. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version of this article.)
onto the trailing hemisphere, resulting in the bullseye pattern
centered on the trailing apex (Fig. 8, upper left). This can be
understood using Eq. (1), with v_ almost purely in the corotation
direction, resulting in a cosðyÞ flux pattern, where y is the angular
distance from the trailing hemisphere apex.
Hendrix et al. (2011) focused on this cold population, reporting
a similar bullseye pattern in the UV surface reflectance, which we
attributed to absorption by SO2 in the surface. That paper
presented some of these results, with a focus on the sulfur ion
flux, and we concluded that the bulls eye of cold sulfur ion flux is
likely responsible for the SO2 bulls eye. The hot sulfur, in
comparison, has negligible flux, though those ions dominate the
sputtering rate, as concluded by Paranicas et al. (2002) and
Plainaki et al. (2010).
At 103 eV, the ion speed becomes comparable to the corotation
drift speed, though the gyroradii are still small compared to
Europa’s radius. At this energy the ion motion parallel to the
magnetic field lines (oriented North/South) becomes important
for reasons we now describe. With increasing ion speed the
corotation drift remains unchanged3 and so the average ion speed
in the corotation direction remains 76 km s 1. The ion motion
parallel to the magnetic field faces no such restriction though, and
at high energies the ions increasingly reach Europa’s surface from
3
The gradient curvature drift also changes with ion energy, but it is small
compared to the corotation drift for the ions considered here.
the North or South. At 105 eV (Fig. 8), the ion flux peaks at the
poles. At even higher energies the gyroradius is comparable to
Europa’s radius and the ions can access Europa’s surface from all
directions, as seen in the nearly uniform flux of the 106 eV case.
Most of the sputtering in our calculation is due to sulfur ions
with energies on the order of 105 eV, yet the corresponding map
in Fig. 8 does not seem to correspond closely to the sputtering
map in Fig. 7. There are two reasons for this: first, the temperature effect (Eq. (2)) enhances sputtering near the equator, offsetting the poles’ advantage in energetic ion flux. Second, the cold
ion population (mostly O þ ) results in a large sputtering rate near
the trailing apex.
In contrast to the sputtering map, energetic H þ and electron
flux peaks at the poles owing to their extremely small gyroradii
and high speeds (Truscott et al., 2010; see companion paper by
Dalton et al. (2012) in this issue).
4.2. O2 sputtering rate
As ions and electrons erode the surface, ejecting H2O molecules, they also create H2, O2 and other species in a process called
radiolysis. O2 created by radiolysis, then subsequently sputtered
or thermally desorbed from the surface, forms the dominant
component of Europa’s tenuous atmosphere (Johnson et al.,
1982; Hall et al., 1998). Here we provide the first estimated
map of its source rate.
T.A. Cassidy et al. / Planetary and Space Science 77 (2013) 64–73
71
O2 Sputtering Rate (cm-2 s-1)
4x109
0
Fig. 9. Calculated O2 sputtering rate on Europa’s surface. Note that unlike the total sputtering rate (Fig. 7), the O2 sputtering rate is mostly confined to the trailing
hemisphere.
The sputtering yields described above are the total sputtering
yields, in this section we examine the H2 and O2 portion. Teolis et al.
(2010) used published laboratory data to develop an empirical
model of O2 production. We used their yields to estimate the O2
production on Europa’s surface. By stoichiometry, the H2 production
rate is exactly twice that of the O2 sputtering rate. This is the first
such estimate of O2 and H2 production. Estimates of these sputtering
rates usually assume that the O2/H2 portion of the yield is some
fixed fraction of the H2O sputtering rate (see, e.g., Saur et al., 1998
and Smyth and Marconi, 2006). In contrast our O2/H2 fraction
depends on projectile, energy and surface temperature.
Fig. 8 and Table 2 show the O2 sputtering rate. Unlike the total
sputtering rate (Fig. 7), the O2 is created almost entirely on the
trailing hemisphere. This is because low-energy heavy ions (like
oxygen), which bombard the trailing hemisphere in a bullseye
pattern, are effective at producing O2. In contrast, the hot ions
dominate the overall sputtering yield owing to their huge advantage in sputtering yield. Energetic ions still have larger O2 yields,
but not the two-order of magnitude advantage seen in Fig. 3.
Teolis et al. (2010) explained why cold ions have high O2
sputtering yields. Radiolytic production is often parameterized
by a quantity called the ‘G value’, the number of molecules
created per unit energy deposited in a surface by ions or
electrons. The G value for O2 production varies by over 4 orders
of magnitude, but Teolis et al. found that they could parameterize
the available data with a single depth-dependent G value, with
the G value decreasing rapidly with depth. Teolis et al. (2005) had
previously found that O2 is produced in the top few nm near the
ice surface even by ions that penetrate much further.
Low-energy heavy ions (like our ‘cold’ Sn þ and On þ ) do not
penetrate deeply into the ice and so deposit most of their energy
near the surface unlike more penetrating energetic ions, such as
the ‘hot’ sulfur ions responsible for most of the H2O sputtering.
Like the total yield, the O2 yield also has temperature dependence. Based on measurements of O2 sputtering yields at various
temperatures the temperature dependence is given by
YðTÞ ¼ YðT ¼ 0Þð1 þ1000 Exp½0:06 eV=kTÞ
which is similar to the temperature dependence of the total
sputtering yield (Eq. (2)), but more sensitive to temperature
(the only difference is the factor of 1000 replacing 220).
This temperature dependence would result in a factor of 5
difference in O2 production between night and day, except that
Teolis et al. (2005) found that the O2 yields have a delayed
response to changes in temperature. They found that it takes a
flux of about 1 1015 ions cm 2, deposited over any length of
time, before the yield adjusts. At Europa’s trailing apex that
corresponds to about 15 Europan days. For that reason we used
the diurnally-averaged O2 sputtering yield.
The H2O sputtering rate is actually nearly independent of
temperature at the relevant temperatures ( 80–130 K), the
temperature dependence in the total sputtering yield (Eq. (2))
arises mostly from production and sputtering of O2 and H2 (Famá
et al., 2008).
Overall, O2 and H2 sputtering are a minor component of the
total sputtering rate ( 10% by mass). O2 is the dominant species
of the atmosphere, however, because it does not freeze to the
surface like H2O. H2 is also abundant in the atmosphere, but it
quickly escapes Europa’s gravity. Assuming the atmosphere has
the same temperature as Europa’s surface, the O2 is gravitationally bound but the H2 is not (Smyth and Marconi, 2006).
There have been few observations of the O2 atmosphere, which
is detected via UV emission triggered by electron impact. Hansen
et al. (2005) observed these emissions during Cassini’s Jupiter
flyby and found that the trailing hemisphere was much brighter
than the leading hemisphere, in seeming agreement with Fig. 9.
Further, occultations of Europa’s ionosphere found it absent near
the leading apex (Kliore et al., 1997). However, HST observations
did not see the same leading/trailing asymmetry (Hall et al.,
1998). Further, an asymmetric source of O2 will not necessarily
result in an asymmetric atmosphere. Cassidy et al. (2007) found
that the O2 spreads over Europa’s surface during its short lifetime.
Relating the O2 source rate to these observations requires treatment of the exciting electrons, and thus treatment of the
magnetosphere/moon interaction (e.g., Saur et al., 1998).
H2 is produced along with the O2 and at twice the rate,
2 1026 s 1. Assuming the H2 has the same temperature as
Europa’s surface, it quickly escapes Europa’s gravity to populate
a ‘cloud’ or ‘torus’ at Europa’s orbit. Such a cloud was inferred by
Mauk et al. and Lagg et al. (2003), who could not definitively
identify the species responsible. Smyth and Marconi (2006)
concluded that H2 from Europa is likely responsible (the other
possibility being atomic O from Europa’s O2 atmosphere). They
were not able to make a careful estimate of the H2 sputtering rate,
but used a value about 10 times larger than we report here,
1.9 1027 s 1.
4.3. Comparison of regolith grain size and sputtering rate
Clark et al. (1983) concluded that regolith grain sizes on the
icy Galilean satellites (i.e., not including Io) are controlled by the
sputtering rate, with larger sputtering rates producing larger
grains. Though all grains are eroded at the same rate (per unit
area), small grains lose mass faster than large grains due to their
larger surface to volume ratio. That mechanism does not create
large grains, simply destroys the smaller ones. In addition to
destroying small grains, the sputtering may also help to grow
large grains as water molecules may be more likely to adsorb onto
72
T.A. Cassidy et al. / Planetary and Space Science 77 (2013) 64–73
water ice grain sizes is necessarily incomplete, and more work
remains to be done. The present results nonetheless contribute to
an improved understanding of the exogenic influences on Europa’s
surface properties and composition. It will be of interest to see
how well the correlation holds up in the future, as additional NIMS
observations at diverse locations are analyzed.
Sputtering represents only one of several important exogenic
influences on Europa. In a companion paper (Dalton et al. (2012),
this issue) we explore the energy deposition into Europa’s surface
as a function of location due to both energetic ions and to
electrons. The interplay of sputtering, energy deposition, endogenic chemistry, and implanted exogenic species makes the study
of Europa’s surface properties and surface chemistry both challenging and rewarding.
6. Conclusion
Fig. 10. Correlation between calculated sputtering rate and estimated grain size.
large grains, known to happen experimentally when the eroding
agent is evaporation (Clark et al., 1983).
We test the idea that sputtering results in larger grains by
looking at the correlation between our calculated sputtering rates
and observed H2O ice grain size. The observations and grain sizes
are summarized in Table 1. These same locations are marked on
the sputtering map in Fig. 7.
Fig. 10 shows the correlation between sputtering rate and
mean grain size from Table 1. The Pearson correlation coefficient
is 0.97. While the existence of a relationship linking H2O ice grain
size with sputtering rate is strongly suggested, correlation does
not demonstrate causation. Caution is advisable, as our physical
description of potentially relevant factors is clearly incomplete;
we have not modeled all processes that might contribute to grain
growth.
Another possibility is that some other exogenic process is
responsible for the trend. Interplanetary impactors, for example,
are expected to hit the leading hemisphere at much larger rates
(Zahnle et al., 1998), and so might explain the small grain sizes
found on the leading hemisphere. However such a process would
be unlikely to explain the nearly identical grain sizes at the
widely separated leading hemisphere observation points (indicated by the asterisk, triangle, and plus signs) along with a strong
trend on the trailing hemisphere (circle, diamond).
The apparently minor difference in mean water ice grain sizes
found for the leading side polar (triangle, 80 mm) and leading side
apex (asterisk, 55 mm) locations may be explained by the difference in sputtering rates obtained for the two locations (note once
again the higher sputtering rate shown for south polar regions in
Fig. 7).
The data point for G1ENNHILAT01 (diamond) corresponds to a
different type of terrain (chaos rather than ridged plains). It is
thus possible that the agreement of this point with the general
trend is fortuitous. However, if we remove this point from the
sample, the correlation remains high at 0.96. Perhaps the most
that we can say is that the selection of large-grained ice for this
location by the linear mixture model algorithms is a robust result
that appears consistent with the trend established by the bright
plains observations.
5. Discussion
Our investigation has uncovered a relationship linking the
sputtering rate at Europa with the distribution of water ice grain
sizes found at widely-spaced locations. Our physical parameterization of all potentially relevant processes affecting surface
We performed the first spatially-dependent calculation of
Europa’s sputtering rate using up-to-date plasma parameters
and sputtering yields. We conclude, as others have previously,
that hot sulfur ions are responsible for the majority of the
sputtering. We performed the first detailed estimate of the O2
and H2 sputtering yields and found that cold heavy ions (sulfur
and oxygen) are responsible for most of the O2 and H2 production.
O2 and H2 are a minor component of the total sputtering rate
( 20% by mass), but of the order of magnitude necessary to
explain the observed O2 atmosphere and H2 torus.
We found that the connection between ion flux and sputtering
rate is complicated. Though most of the sputtering is done by
energetic sulfur ions whose flux peaks at the poles, the overall
sputtering map shows the long-assumed leading/trailing asymmetry (e.g., Tiscareno and Geissler, 2003), with the sputtering rate
peaked on the trailing apex. One finding from our approach is that
cold oxygen can be competitive with energetic heavy ions in
sputtering Europa’s surface on the trailing hemisphere.
We found that experimental sputtering data is best fit by a
combination of two analytic sputtering formulas. Below 105 eV,
the formula from Famá et al. (2008) best fits the data. Above
105 eV the formula from Johnson et al. (2009) best fits the data.
We reported new spectral fits to NIMS observations of the trailing
hemisphere apex. The estimated surface composition shows a
strong magnetospheric influence, including high concentration of
sulfuric acid hydrate and radiation-resistant hydrated salts. We
concluded that the sputtering rate is highly correlated with
regolith grain size as suggested by Clark et al. (1983), but did
not quantify the processes responsible. The NIMS observations
and fits are described by Dalton et al. (2012) (this issue) in more
detail.
Acknowledgments
TAC would like to thank Michael Kokorowski for useful
discussions and for pointing out the Roederer (1970) reference.
TAC and REJ thank Min Liu, at the University of Virginia, for help
in compiling the sputtering yield data. Part of this research was
carried out at the Jet Propulsion Laboratory, California Institute of
Technology, under a contract with the National Aeronautics and
Space Administration.
References
Bagenal, F., 1994. Empirical model of the IO plasma torus: voyager measurements.
Journal of Geophysical Research 99, 11043–11062.
Carlson, R.W., Johnson, R.E., Anderson, M.S., 1999. Sulfuric acid on Europa and the
radiolytic sulfur cycle. Science 286, 97–99.
T.A. Cassidy et al. / Planetary and Space Science 77 (2013) 64–73
Cassidy, T.A., Johnson, R.E., 2005. Monte Carlo model of sputtering and other
ejection processes within a regolith. Icarus 176, 499–507.
Cassidy, T.A., Johnson, R.E., Geissler, P.E., Leblanc, F., 2008. Simulation of Na D
emission near Europa during eclipse. Journal of Geophysical Research 113,
E02005.
Cassidy, T.A., Johnson, R.E., McGrath, M.A., Wong, M.C., Cooper, J.F., 2007. The
spatial morphology of Europa’s near-surface O2 atmosphere. Icarus 191,
755–764.
Clark, R.N., Fanale, F.P., Zent, A.P., 1983. Frost grain size metamorphism: implications for remote sensing of planetary surfaces. Icarus 56, 233–245.
Cooper, J.F., Johnson, R.E., Mauk, B.H., Garrett, H.B., Gehrels, N., 2001. Energetic ion
and electron irradiation of the icy galilean satellites. Icarus 149, 133–159.
Dalton III, J.B., Shirley, J.H., Cassidy, T.A., Paranicas, C.P., 2012. Exogenic and
endogenic controls on the radiolytic sulfur cycle at Europa. Planetary and
Space Science.
Dalton, J.B., 2007. Linear mixture modeling of Europa’s non-ice material based on
cryogenic laboratory spectroscopy. Geophysical Research Letters 34, L21205.
Domingue, D.L., Verbiscer, A., 1997. Re-analysis of the solar phase curves of the icy
Galilean satellites. Icarus 128, 49–74.
Famá, M., Shi, J., Baragiola, R.A., 2008. Sputtering of ice by low-energy ions. Surface
Science 602, 156–161.
Hall, D.T., Feldman, P.D., McGrath, M.A., Strobel, D.F., 1998. The far-ultraviolet
oxygen airglow of Europa and Ganymede. The Astrophysical Journal 499, 475.
Hansen, C.J., Shemansky, D.E., Hendrix, A.R., 2005. Cassini UVIS observations of
Europa’s oxygen atmosphere and torus. Icarus 176, 305–315.
Hendrix, A.R., Cassidy, T.A., Johnson, R.E., Paranicas, C., Carlson, R.W., 2011.
Europa’s disk-resolved ultraviolet spectra: relationships with plasma flux
and surface terrains. Icarus 212, 736–743.
Roederer, J.G., 1970. Dynamics of Geomagnetically Trapped Radiation. SpringerVerlag, New York.
Johnson, R.E., Burger, M.H., Cassidy, T.A., Leblanc, F., Marconi, M., Smyth, W.H.,
2009. Composition and Detection of Europa’s Sputter-Induced Atmosphere. In:
Pappalardo, R.T., McKinnon, W.B., Khurana, K.K. (Eds.), Europa. University of
Arizona Press, Tucson.
Johnson, R.E., Lanzerotti, L.J., Brown, W.L., 1982. Planetary applications of ion
induced erosion of condensed-gas frosts. Nuclear Instruments and Methods in
Physics Research 198, 147–157.
Kivelson, M.G., Khurana, K.K., Volwerk, M., 2009. Europa’s Interaction with the
Jovian Magnetosphere. In: Pappalardo, R.T., McKinnon, W.B., Khurana, K.K.
(Eds.), Europa. University of Arizona Press, Tucson.
Lagg, A., Krupp, N., Woch, J., Williams, D.J., 2003. In-situ observations of a neutral
gas torus at Europa. Geophysical Research Letters 30, 1556.
Mauk, B.H., Mitchell, D.G., McEntire, R.W., Paranicas, C.P., Roelof, E.C., Williams,
D.J., Krimigis, S.M., Lagg, A., 2004. Energetic ion characteristics and neutral gas
interactions in Jupiter’s magnetosphere. Journal of Geophysical Research 109,
A09S12.
McCord, T.B., Orlando, T.M., Teeter, G., Hansen, G.B., Sieger, M.T., Petrik, N.G., Van
Keulen, L., 2001. Thermal and radiation stability of the hydrated salt minerals
epsomite, mirabilite, and natron under Europa environmental conditions.
Journal of Geophysical Research 106, 3311–3319.
Paranicas, C., Cooper, J.F., Garrett, H.B., Johnson, R.E., Sturner, S.J., 2009. Europa’s
Radiation Environment and Its Effects on the Surface. In: Pappalardo, R.T.,
73
McKinnon, K.K., Khurana, K.K. (Eds.), Europa. University of Arizona Press,
Tucson.
Paranicas, C., Ratliff, J.M., Mauk, B.H., Cohen, C., Johnson, R.E., 2002. The ion
environment near Europa and its role in surface energetics. Geophysical
Research Letters 29, 1074.
Plainaki, C., Milillo, A., Mura, A., Orsini, S., Cassidy, T., 2010. Neutral particle release
from Europa’s surface. Icarus 210, 385–395.
Pospieszalska, M.K., Johnson, R.E., 1989. Magnetospheric ion bombardment
profiles of satellites: Europa and Dione. Icarus 78, 1–13.
Johnson, R. E., 1990. Energetic Charged-Particle Interactions with Atmospheres
and Surfaces.
Rodriguez-Nieva, J.F., Bringa, E.M., Cassidy, T.A., Johnson, R.E., Caro, A., Fama, M.,
Loeffler, M.J., Baragiola, R.A., Farkas, D., 2011. Sputtering from a Porous
material by penetrating ions. The Astrophysical Journal Letters 743, L5.
Saur, J., Strobel, D.F., Neubauer, F.M., 1998. Interaction of the Jovian magnetosphere with Europa: Constraints on the neutral atmosphere. Journal of
Geophysical Research 103, 19947–19962.
Shirley, J.H., Dalton III, J.B., Prockter, L.M., Kamp, L.W., 2010. Europa’s ridged plains
and smooth low albedo plains: distinctive compositions and compositional
gradients at the leading side–trailing side boundary. Icarus 210, 358–384.
Smyth, W.H., Marconi, M.L., 2006. Europa’s atmosphere, gas tori, and magnetospheric implications. Icarus 181, 510–526.
Spencer, J.R., Lebofsky, L.A., Sykes, M.V., 1989. Systematic biases in radiometric
diameter determinations. Icarus 78, 337–354.
Spencer, J.R., Tamppari, L.K., Martin, T.Z., Travis, L.D., 1999. Temperatures on
Europa from Galileo photopolarimeter-radiometer: nighttime thermal anomalies. Science (New York, NY) 284, 1514–1516.
Teolis, B.D., Jones, G.H., Miles, P.F., Tokar, R.L., Magee, B.A., Waite, J.H., Roussos, E.,
Young, D.T., Crary, F.J., Coates, A.J., Johnson, R.E., Tseng, W.-.L, Baragiola, R.A.,
2010. Cassini finds an oxygen–carbon dioxide atmosphere at Saturn’s icy
moon rhea. Science (New York, NY) 330, 1813–1815.
Teolis, B.D., Vidal, R.A., Shi, J., Baragiola, R.A., 2005. Mechanisms of O2 sputtering
from water ice by keV ions. Physical Review B 72, 245422.
Thomsen, M.F., Van Allen, J.A., 1980. Motion of Trapped electrons and protons in
Saturn’s inner magnetosphere. Journal of Geophysical Research 85,
5831–5834.
Tiscareno, M.S., Geissler, P.E., 2003. Can redistribution of material by sputtering
explain the hemispheric dichotomy of europa? Icarus 161, 90–101.
Truscott, P., Heynderickx, D., Nartallo, R., Lei, F., Sicard-Piet, A., 2010. Application of
PLANETOCOSMICS to Simulate the Radiation Environment at the Galilean
Moons. Presentation at the European Planetary Space Conference.
Paterson, W.R., Frank, L.A., Ackerson, K.L., 1999. Galileo plasma observations at
Europa: ion energy spectra and moments. Journal of Geophysical Research
104, 22779–22792.
Williams, D.J., Mauk, B., 1997. Pitch angle diffusion at Jupiter’s moon Ganymede.
Journal of Geophysical Research 102, 24283–24287.
Zahnle, K., Dones, L., Levison, H.F., 1998. Cratering Rates on the Galilean Satellites.
Icarus 136, 202–222.
Zimmer, C., Khurana, K.K., Kivelson, M.G., 2000. Subsurface oceans on Europa and
Callisto: constraints from Galileo magnetometer observations. Icarus 147,
329–347.